Nemotron 3.5 Embed 1B is an open-source embedding model created by NVIDIA. It takes text and image inputs to produce vector embeddings, designed for multimodal and multilingual retrieval tasks such as question-answering over document collections.
Input
Output
Context
262K
Max Output
n/a
Parameters
1.6B
Input Modalities
Output Modalities
None the catalog can show from stated facts or measured scores.
| Precision | Weights | Smallest GPU that fits |
|---|---|---|
| FP16 / BF16 | 2.90 GB | 1x NVIDIA T4 (16 GB) |
| INT8 | 1.45 GB | 1x NVIDIA T4 (16 GB) |
| INT4 | 0.72 GB | 1x NVIDIA T4 (16 GB) |
Computed by Inferbase from the 1,554,710,528 parameters in the creator's Hugging Face repository; weights only, as serving adds the framework and the input being processed.
Public resources
Specification sources
As of Nov 15, 2026.
Answered from the facts on this page, with their sources.
Nemotron 3.5 Embed 1B has a context window of 262,144 tokens (source: Creator or provider API). A provider may serve less; Inferbase routes on what each provider serves.
Yes. Nemotron 3.5 Embed 1B is an open-weight model under the Other license. You can download and self-host it within the license terms.
Nemotron 3.5 Embed 1B has 1,554,710,528 parameters (1.6B), as stated in the creator's Hugging Face repository.
Strengths: Supports 33 languages. Each is computed from the model's stated facts and benchmark results.
Nemotron 3.5 Embed 1B accepts Text and Image and produces Embedding. The Capabilities card lists the API features it supports, such as function calling and structured output.
No benchmark source has published a result for Nemotron 3.5 Embed 1B yet, so it is shown as unmeasured, not estimated.
Yes. Nemotron 3.5 Embed 1B is available on Inferbase Managed through the OpenAI-compatible API as nvidia-nemotron-3-5-embed-1b. Request it by name, or request auto and routing picks it when it fits the request best.
The weights take 2.9 GB at FP16 / BF16, 1.4 GB at INT8, 0.7 GB at INT4, computed from the 1,554,710,528 parameters its creator publishes. Serving adds the framework and the input being processed.
The model card of Nemotron 3.5 Embed 1B lists 33 languages: English, Arabic, Assamese, Bangla, Bulgarian, Chinese, Danish, Dutch, Finnish, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Malay, Marathi, Nepali, Norwegian, Persian, Portuguese, Romanian, Russian, Spanish, Swahili, Swedish, Tamil, Telugu, Thai, Ukrainian, Urdu and Vietnamese. That is what NVIDIA states; the catalog does not test language quality.
Start on the free plan with 5,000 requests a month, no card required.